Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752440AbdHPVp4 (ORCPT ); Wed, 16 Aug 2017 17:45:56 -0400 Received: from mail-oi0-f43.google.com ([209.85.218.43]:35889 "EHLO mail-oi0-f43.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752298AbdHPVpz (ORCPT ); Wed, 16 Aug 2017 17:45:55 -0400 MIME-Version: 1.0 In-Reply-To: References: <20170816171211.4021-1-christian.brauner@ubuntu.com> <20170816194805.hnof3aqiqykwki7p@gmail.com> From: Linus Torvalds Date: Wed, 16 Aug 2017 14:45:54 -0700 X-Google-Sender-Auth: VqSsQSgI-Eb8Wd8ni7YAAhkfe78 Message-ID: Subject: Re: [PATCH 0/1] devpts: use dynamic_dname() to generate proc name To: Christian Brauner Cc: Christian Brauner , Linux Kernel Mailing List , "Serge E. Hallyn" , Al Viro Content-Type: text/plain; charset="UTF-8" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1217 Lines: 26 On Wed, Aug 16, 2017 at 2:37 PM, Christian Brauner wrote: >> And Christian, if you can beat on this, that would be good. > > Yes, I can pound on this nicely with liblxc. We have patch > ( https://github.com/lxc/lxc/pull/1728 ) up for review that > allocates pty fds from private devpts mounts in different namespaces > and sends those fds around between different namespaces. Good. Testing that this works with different pts filesystems in different places is exactly the kind of thing I'd like to see. I only tested with my single pts filesystem that is mounted at /dev/pts, and making sure it works when there are multiple mounts and in different places is exactly the kind of testing this should get. For example, if some namespace has it's pty's in _its_ /dev/pts/ hierarchy, and you then pass such a pty to somebody else that either doesn't have that pts mount at all, or has it visible somewhere entirely different, the result should now be something else than that "/dev/pts/n" path. But it would be good to just test this in general too, and make sure I didn't screw up some reference count or something. The patch *looks* obviously correct, but ... Linus